Glossary
Dictionary for the administrator.
Who is this page for?
When working with the platform, you come across the terms Kubernetes, CAPI and Shturval-specific terms. This page contains brief definitions for quick orientation.
Installation and mirror
| Term | Definition |
|---|---|
| shtil | Shturval platform installation utility. Launches the installation wizard in the browser or performs installation via the console (YAML, flags). |
| Machine shtil | The virtual or physical machine on which the shtil initial deployment utility runs. |
| stm (Shturval Mirror) | A utility for working with a mirror of images and packages. Initializing the mirror, downloading the bundle, updating, deleting releases. Included in the bundle. |
| stc | A utility for preparing virtual machine templates for providers (vSphere, oVirt, etc.). |
| Open loop | An installation and upgrade scenario in which the cluster obtains images and packages from external repositories. Does not require a local mirror. |
| Closed loop | Scenario without nodes access to the Internet. Requires a local mirror, bundle and Enterprise license. |
| Conditionally closed loop | A closed loop option in which cluster nodes do not have access to the Internet, but the mirror host can download bundles from external sources. |
| Bootstrap-cluster | A temporary installation cluster that shtil uses during the deployment phase of a permanent management cluster. |
| Bundle | tar archive with images and packages for a specific version of the platform. Used in a closed loop where there is no access to the Internet. |
| Mirror | Local repository of images and packages (Nexus, Nginx). Nodes download the distribution from the mirror when installing in a closed or conditionally closed loop. |
| Nexus | Artifact storage system (images, packages). Built into stm or external Nexus can be used. |
| CE (Community Edition) | Community platform license. Suitable for open loop. |
| Enterprise | Commercial edition of the platform. Required for closed loop, mirror and DR. |
| Enterprise ML | Commercial edition with capabilities for ML scripts. Inherits the limitations and capabilities of Enterprise unless otherwise noted in the documentation. |
| NTP (Network Time Protocol) | Time synchronization protocol. Time synchronization is required on all cluster nodes. |
| SELinux | Linux security subsystem. Must be disabled or permissive on nodes. |
| ShturvalUpdateChannel | Custom resource (API group update.shturval.tech). Determines available versions for update. After updating the mirror, restarting the resource speeds up the appearance of the new version. |
| Step-CA | Certificate issuing tool (alternative to ACME). Can be used for corporate mirror certificates. |
| CRI (Container Runtime Interface) | Interface Kubernetes to container runtime (containerd, cri-o). Used when configuring nodes. |
| GPG (GnuPG) | Encryption tool. Mirror apt repositories require a pair of GPG keys to sign packages. |
| Service UZ (account) | Credentials for connecting the platform to vCenter, oVirt, OpenStack, CSI. When changing the password in the infrastructure, they need to be updated in the platform. |
Clusters and nodes
| Term | Definition |
|---|---|
| Management cluster (management cluster) | The cluster on which the Shturval platform runs. Manages client clusters, stores providers, access rights, metrics. |
| Client cluster (workload cluster) | The Kubernetes cluster created by the platform to host applications. Subordinate to the management cluster. |
| Stand-alone | The management cluster operating mode, in which it is also used for workloads - applications are placed in the same cluster as the platform, without creating client clusters. Suitable for small scenarios. It differs from the normal mode in the absence of a separate workload cluster. More details: Operating modes, open loop requirements, closed loop requirements. |
| Microcluster | Minimum platform configuration on one Control Plane node (6 vCPU, 8 GiB RAM, 80 GiB disk). Suitable for testing, demonstrations, pilot projects. Installation takes 30–60 minutes. More details: Quick microcluster installation. |
| Control Plane (CP) | Cluster management nodes Kubernetes. Launch the API server, etcd, scheduler, controller-manager. Minimum 1, for fault tolerance - 3. |
| Worker | Nodes for hosting applications (pods). They run kubelet, CNI, CSI. |
| Infra Worker | A special group of Worker nodes for system services (monitoring, logging, Ingress). |
| Shturval v2 | Infrastructure provider for prepared bare metal or virtual hosts connected via SSH. It does not use VM templates. |
| Hybrid Cluster | Client cluster in which the main providercreates basic nodes, and additional Worker groups are connected via Shturval v2, for example, for physical nodes with GPUs. |
| External etcd | A client cluster configuration in which the etcd storage is located separately from the Control Plane nodes. Used in special fault tolerance scenarios. |
Network and addresses
| Term | Definition |
|---|---|
| VIP (Virtual IP) | Virtual IP address for accessing KubeAPI or Ingress. It is not permanently assigned to one node and fails over to another node. The platform is managed by the shturval-vip and shturval-vip-provider components based on Kube-VIP. |
| FQDN (Fully Qualified Domain Name) | Fully qualified domain name. For example, shturval.clustername.ip-10-11-12-13.shturval.link is the FQDN of the GUI when using VIP. |
| KubeAPI | Access point to the API server Kubernetes. When creating a cluster, it is specified as VIP or FQDN and is usually accessible on TCP port 6443. |
| Ingress | Entry point for HTTP/HTTPS traffic to applications in the cluster. Can use VIP, FQDN or external balancer. |
| IP Pool | The range of addresses from which the platform assigns VIPs for KubeAPI and Ingress when creating a cluster. Not used by all infrastructure providers. |
| External balancer | A load balancer outside the cluster, through which you can organize access to KubeAPI or Ingress instead of internal VIPs. For some providers it is mandatory. |
| CNI (Container Network Interface) | Kubernetes interface for connecting network plugins. In the platform, the network subsystem is managed by the shturval-networking service. |
| Cilium | CNI plugin used by the platform for network connectivity of clusters and network policies. |
CoreHA (shturval-coreha) | Component for working with dynamic HAProxy backends via DNS. Used in scenarios with an external balancer. |
Providers and infrastructure
| Term | Definition |
|---|---|
| Infrastructure Provider | Virtualization, cloud or physical hosts, where the platform creates or connects client cluster nodes: VMware vSphere, oVirt, OpenStack, Shturval v2, Yandex Cloud, vCloud Director, Basis Dynamix, etc. |
| CAPI Infrastructure Provider | A component in a management cluster that implements the Cluster API contract for a specific infrastructure, such as shturval-capvs for vSphere or shturval-capov for oVirt. |
| Provider Instance | Configured connection to a specific environment (one vCenter, one oVirt Engine, etc.). There can be multiple instances of one provider type. |
| VM Template | An image of a virtual machine with a prepared OS. The platform clones it when creating cluster nodes. Prepared by the stc utility. |
| CAPI (Cluster API) | The Kubernetes standard for declarative lifecycle management of clusters and nodes. In the platform, CAPI resources are coordinated in the management cluster. |
| CAPVS | Infrastructure CAPI platform provider for VMware vSphere (shturval-capvs). |
| CAPOV | Infrastructure CAPI platform provider for oVirt and compatible environments (shturval-capov). |
| CAPOS | Infrastructure CAPI platform provider for OpenStack and compatible clouds (shturval-capos). |
| CAPY | Infrastructure CAPI platform provider for Yandex Cloud (shturval-capy). |
| CAPBD | Infrastructure CAPI platform provider for Basis Dynamix (shturval-capbd). |
| CAPVCD | Infrastructure CAPI platform provider for VMware vCloud Director (shturval-capvcd). |
| CAPSM | Infrastructure CAPI platform provider for Shturval v2 (shturval-capsm). |
| CSI (Container Storage Interface) | Kubernetes interface for connecting storages. The platform is used to integrate provider storage and local CSI storage. |
| MachineDeployment | The Cluster API resource that describes a group of Worker nodes and controls their creation, scaling and updating. |
| MachineHealthCheck | Cluster API resource for checking the state of machines and automatically recovering unavailable nodes. |
Configuration and services
| Term | Definition |
|---|---|
| ShturvalServiceConfig (SSC) | Description of the platform service (chart, parameters). An analogue of a service “template”. |
| ShturvalServicePatch (SSP) | Patch for the service - redefining parameters via custom values. Allows you to customize the service without changing the basic ShturvalServiceConfig (SSC). |
| ClusterProfile | Cluster profile is a set of settings (services, NCI, network policies) used when creating a cluster. |
| NodeConfigItem (NCI) | Configuration applied to cluster nodes (for example, sysctl settings, kernel modules). |
| ClusterConfig | Custom management cluster configuration resource. Stores parameters like FQDN, registry and network endpoints. |
| custom values | A set of Helm chart values that are set via ShturvalServicePatch to change platform service parameters. |
Aboutplatform update
| Term | Definition |
|---|---|
| Shturval Updater | An update module (shturval-update) that manages platform and Kubernetes versions and updates cluster system components. |
| ShturvalUpdate | Custom update task resource. Sets the target version and displays the process status. |
| ShturvalUpdateSchedule | Custom resource for delayed update according to schedule. Describes the schedule and window in which the update can be performed. |
Access and security
| Term | Definition |
|---|---|
| LDAP (Lightweight Directory Access Protocol) | Protocol for accessing the corporate directory of users and groups. Used to authenticate and search for users. |
| OIDC (OpenID Connect) | Federated authentication protocol through an external Identity Provider, such as Keycloak or Blitz. |
| IdP (Identity Provider) | An external identity provider that verifies the user’s identity and provides login information to the platform. |
| Set of access (Permission) | A set of rights (platform, cluster, role in Argo CD, etc.) assigned to a user or group. In the API it is represented by the resource Permission from the group permissions.shturval.tech. |
| Platform Permissions | Rights to platform objects (creating clusters, managing providers, assigning rights, etc.). |
| Cluster Permissions | RBAC rights in Kubernetes (API groups, resources, verbs). |
| Scope | A system resource that defines the scope of application of a set of accesses: platform, cluster, tenant or namespace. |
| UserRole | A resource for assigning rights to a specific user. Refers to one or more access sets and target scope. |
| GroupRole | A resource for assigning rights to a group of users. All group members receive the rights specified through GroupRole. |
| Tenant | An object for grouping clusters or namespaces for assigning rights and controlling access. |
Backup and DR
| Term | Definition |
|---|---|
| Velero | A tool for backing up and restoring Kubernetes clusters. Used in the backup platform. |
| DR (Disaster Recovery) | Disaster tolerance - a configuration with a backup management cluster for switching in the event of failure of the main one. Available in Enterprise and Enterprise ML. |
Monitoring and logging
| Term | Definition |
|---|---|
| Victoria Metrics | System for storing and querying metrics (analogous to Prometheus). Used for cluster metrics. |
| VictoriaLogs (VLogs) | Log storage system. Logs are collected by Vector and queried via VictoriaLogs. |
| Vector | Log collection and routing agent used by the shturval-log-collector service. |
| ClusterVectorPipeline | A custom resource for setting up a pipeline for collecting and routing logs in a cluster. |
| Grafana | Metric visualization tool. Dashboards come with the platform by default. |
Glossaries of other sections
| Section | Description |
|---|---|
| Cluster Administrator Glossary | Terms on cluster management, storage, nodes, monitoring |
| Developer’s Glossary | Application Delivery Terms, GitOps, Argo CD |
| Security Glossary | Terms on policies, scanning, auditing |